[x265] [PATCH] asm code for blockcopy_sp, 6x8
Steve Borho
steve at borho.org
Wed Nov 6 22:34:26 CET 2013
On Wed, Nov 6, 2013 at 10:08 AM, Praveen Tiwari <
praveen at multicorewareinc.com> wrote:
> Fixed.
>
> Regards,
> Praveen Tiwari
>
>
>
> On Wed, Nov 6, 2013 at 8:09 PM, chen <chenm003 at 163.com> wrote:
>
>> >+ movd [r0 + 2 * r1], m3
>> >+ pextrw r6, m3, 2
>> >+ mov [r0 + 2 * r1 + 4], r6w
>> SSE4.1 support below:
>> pextrw [r0 + 2 * r1 + 4], m3, 2
>>
>
When you change the function signature to SSE4, you also need to move the
pointer setups to the SSE4 portion of asm-primitives.cpp, else you defeat
the whole concept of runtime CPU arch identification
--
Steve Borho
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20131106/18e1ec64/attachment.html>
More information about the x265-devel
mailing list